CVE-2026-98859

In the Linux kernel, the following vulnerability has been resolved:

HID: multitouch: stop the release timer from being rearmed on remove

mt remove() quiesces the sticky-finger timer before stopping the hardware:

timer delete sync(&td->release timer);

sysfs remove group(&hdev->dev.kobj, &mt attribute group); hid hw stop(hdev);

timer delete sync() waits for a running callback and dequeues the timer, but it does not stop the timer from being armed again. The transport is still delivering reports at that point, and the report path rearms it:

if (app->quirks & MT QUIRK STICKY FINGERS) { if (td->mt io flags & MT IO SLOTS MASK) mod timer(&td->release timer, jiffies + msecs to jiffies(100));

A report that arrives after timer delete sync() has returned therefore leaves the timer queued. td is allocated with devm kzalloc() against hdev->dev, so it is freed when the driver is unbound, after mt remove() returns. When the timer fires afterwards, mt expired timeout() dereferences the freed td:

struct mt device *td = timer container of(td, t, release timer); struct hid device *hdev = td->hdev;

if (test and set bit lock(MT IO FLAGS RUNNING, &td->mt io flags))

Simply moving the teardown after hid hw stop() does not fix this on its own, because mt expired timeout() calls mt release contacts(), which walks hdev->inputs; the timer still has to be quiesced before hid hw stop() tears the input devices down.

Use timer shutdown sync() instead, which additionally makes any later mod timer() a no-op, so neither ordering constraint has to be traded off against the other. This is the final-teardown pattern the function was introduced for, and hid-wiimote already uses it for the same reason.